4 min read

System Architecture

AIPI turns the Vantage platform into a fully headless Backend as a Service (BaaS). Your systems, apps, and software connect to Intuidy as their backend — no servers to build, no infrastructure to manage. This page explains how the layers fit together and how requests flow through the system.


Layered Architecture

AIPI follows a layered BaaS design where your application consumes backend services without managing any of the underlying infrastructure:

Loading diagram…

Layer Responsibilities

LayerWhat It Does
Your ApplicationThe frontend you build — handles UI, user interactions, and local state
AIPI LayerAuthenticates requests, routes them to the correct service, enforces rate limits and permissions
Vantage Core ServicesProcesses business logic — executes workflows, runs AI agents, manages integrations, generates dashboard data
Data LayerStores and retrieves data from connected databases, manages file storage, and handles event streaming

Request Lifecycle

When your application makes an AIPI request, it follows this path:

  1. Your app sends an authenticated request
  2. AIPI validates credentials and permissions
  3. Request is routed to the appropriate Vantage service
  4. The service processes the request (query data, execute workflow, call AI, etc.)
  5. Response is formatted and returned to your app

Authentication Flow

Every AIPI request requires authentication. The platform supports:

MethodHow It WorksBest For
API KeyInclude your key in the request headerServer-to-server communication, backend services
Session TokenAuthenticate users through the Vantage login flowUser-facing applications where individuals sign in

Multi-Tenant Isolation

AIPI inherits the Vantage multi-tenant architecture. Data is isolated at the company level:


Core Service Connections

AIPI connects your application to every major Vantage subsystem:

Loading diagram…
ServiceWhat AIPI Exposes
Workflow EngineTrigger workflows, pass input data, retrieve output results, monitor execution status
Dashboard EngineRead dashboard data, push data from external sources into dashboard tiles
AI SystemSend prompts to configured AI models, deploy and manage AI agents
IntegrationsAccess connected services (email, calendar, storage, CRM, etc.) through a unified interface
Database LayerExecute queries, insert/update/delete records, manage schemas
SecurityManage users, roles, and permissions programmatically

Data Flow Patterns

Read Pattern

Your app requests data → AIPI authenticates → Vantage queries the data source → structured response returned.

Write Pattern

Your app sends data → AIPI validates and authenticates → Vantage writes to the target database or service → confirmation returned.

Trigger Pattern

Your app fires an event → AIPI routes to the Workflow Engine → workflow executes (can involve multiple services, AI analysis, notifications) → final result returned or pushed asynchronously.

Agent Pattern

Your app sends a message to an AI agent → AIPI routes to the deployed agent → agent processes using its skills, data resources, and tools → response streamed back.


Next Steps